Guide Dogs Maximising the impact of Raffle appeals

Guide Dogs Maximising the impact of Raffle appeals Kate Parker Senior Insight Analyst

Guide Dogs Raffles - Overview § One of our longest running appeals § Quarterly appeals § Sent over 3 m warm mailings in 2011 § To 880 k warm supporters in 2011 § Raised £ 5 m in 2011 (warm and cold)

The Creative

The Problem… § Mass communication § Not targeted § Great way to keep in touch with supporters § Potential income slipping through

The Proposal… § Identify max profit line and mail minimum § Send more books to supporters who are most likely to buy more tickets § Use data to personalise the letter to make the supporter feel more engaged with the cause

Part 1 - Modelling § Analyse past campaigns : – – – Who we mailed Who responded What differentiates them from the base? RFV variables & demographics Identify max profit line § Use the model to select lower volumes to increase RR% and ROI.

Modelling – Profiling responders

Modelling – Decision Tree § afd

Modelling – Gains Chart

Modelling - Results

Part 2 – Multibook Strategy § Increase average gift by sending more tickets § Which supporters have the propensity to buy more? § How many more tickets do we send them? § Who are the ‘stubborn’ supporters?

Multibook Strategy - Planning § Which audience? – Non-rafflers were tested with multiple books in Summer & Winter 2010 – Those sent multiple books underperformed § Are we potentially putting off non-rafflers by mailing them too many books? § Focus on best responders to increase average gift

Multibook Strategy – The strategy

Multibook Strategy – The strategy

Multibook Strategy – The results § Response rate decreased for almost all test data Spring 2011 Summer 2011 Autumn 2011 Winter 2011 New to 2 NO TEST -9. 35% -8. 24% -5. 52% 1 to 2 -6. 42% -10. 98% -14. 04% -18. 81% 2 to 3 2. 02% -2. 48% -0. 73% -12. 81% § However, average gift was much higher across the board. Spring 2011 Summer 2011 Autumn 2011 Winter 2011 New to 2 NO TEST 27. 41% 25. 64% 25. 72% 1 to 2 22. 08% 26. 60% 19. 18% 17. 37% 2 to 3 7. 66% 6. 94% 5. 58% 12. 45%

Multibook Strategy – The results § Need to examine ROI to determine effectiveness § Test segments performed with mixed results § Consistently positive results for New Rafflers – increased ROI for those sent more books § Increasing books for existing supporters looked positive at first, but then decreased ROI Spring 2011 Summer 2011 Autumn 2011 Winter 2011 New to 2 NO TEST 4. 75% 4. 85% 8. 44% 1 to 2 3. 65% 2. 22% -6. 82% -13. 01% 2 to 3 4. 96% -0. 38% -3. 90% -9. 81%

Multibook Strategy – The results § Moving poor responders down meant response rate for other cells dramatically increased

Multibook Strategy - Conclusions § Existing raffle supporters are happy with the level of gift they have given – maintain & monitor § New raffle supporters can be encouraged to buy more without negatively affecting ROI § Moving supporters down in the strategy strengthened results

Maximising Raffles - Conclusions § Multi-faceted approach is best § Data and analysis needs to be robust to accurately predict behaviour § Success is measured in various ways – identifying this from the outset helps your strategy

Any questions?